6 Mr. 0. J. Lodge. Experiment* on [June 4, 



through the long thick wire, the spark-gap F through the capillary 

 water tube of high resistance already mentioned. The A knobs were 

 each 2-34 cm. diameter. The size of the others does not seem to be 

 recorded. 



FIG. 2. 



Separating the machine knobs too far for a spark there, sparks 

 could be got either at C or at F or over the lip of the jar, or in two or 

 three places at once. The lengths were D = 0*72 inch, F = 0'68 inch. 

 Bringing the A knobs nearer together, a distance of 0'57 inch, it went 

 there too. The A spark is the noisiest, then D, and lastly F ; F is in 

 fact quite weak. When it sparks at D it mostly goes at F too, and 

 likewise overflows the lip of the jar, but not always. 



Shorten all the air-gaps so as to avoid overflow, and they spark 

 simultaneously at the following distances : 



A. D. F. 



0-435 0-565 0'575 



Modified the plan of connexions to that shown in fig. 3 ; the second 

 water resistance or " leak " being now introduced merely in order to 

 give the jar the possibility of charging. 



Whenever an A sj^ark occurs, a considerable range is permissible 

 with the others. As to F, it does not matter how short that is made ; 

 it is affected by the others, but has no effect on them. The overflow 

 of jar specially accompanies a spark at D. Frequently sparks occur 

 in all four places at once ; and at times the overflows of jars are 

 violent and numerous, so that, when A and D are both pretty long, 

 flashes fly from cork and wood and almost anything that happens to 

 be in contact with the jar. (The jar stood on a wooden block on an 

 insulating stool : it was principally from this that flashes sprang 

 sometimes.) 



